Open Access Available online http://ccforum.com/content/9/5/R583 R583 Vol 9 No 5 Research Hospital-acquired sinusitis is a common cause of fever of unknown origin in orotracheally intubated critically ill patients Arthur RH van Zanten 1 , J Mark Dixon 2 , Martine D Nipshagen 3 , Remco de Bree 4 , Armand RJ Girbes 5 and Kees H Polderman 6 1 Senior Consultant in Internal Medicine and Intensive Care, Department of Intensive Care, Gelderse Vallei Hospital, Ede, The Netherlands 2 Senior Consultant in Anaesthesiology and Intensive Care, Department of Anesthesiology and Intensive Care, Norfolk and Norwich University Hospital, Norwich, UK 3 Resident in Plastic Surgery, Hospital Hilversum, Hilversum, The Netherlands 4 Professor of Intensive Care Medicine, Department of Intensive Care, VU University Medical Center, Amsterdam, The Netherlands 5 Senior Consultant in Otolaryngology, Department of Otolaryngology/Head and Neck Surgery, VU University Medical Center, Amsterdam, The Netherlands 6 Senior Consultant in Intensive Care, Department of Intensive Care, VU University Medical Center, Amsterdam, The Netherlands Corresponding author: Kees H Polderman, k.polderman@vumc.nl Received: 21 Jun 2005 Revisions requested: 27 Jul 2005 Revisions received: 9 Aug 2005 Accepted: 12 Aug 2005 Published: 13 Sep 2005 Critical Care 2005, 9:R583-R590 (DOI 10.1186/cc3805) This article is online at: http://ccforum.com/content/9/5/R583 © 2005 van Zanten et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/ 2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Abstract Introduction Sinusitis is a well recognised but insufficiently understood complication of critical illness. It has been linked to nasotracheal intubation, but its occurrence after orotracheal intubation is less clear. We studied the incidence of sinusitis in patients with fever of unknown origin (FUO) in our intensive care unit with the aim of establishing a protocol that would be applicable in everyday clinical practice. Methods Sinus X-rays (SXRs) were performed in all patients with fever for which an initial screening (physical examination, microbiological cultures and chest X-ray) revealed no obvious cause. All patients were followed with a predefined protocol, including antral drainage in all patients with abnormal or equivocal results on their SXR. Results Initial screening revealed probable causes of fever in 153 of 351 patients (43.6%). SXRs were taken in the other 198 patients (56.4%); 129 had obvious or equivocal abnormalities. Sinus drainage revealed purulent material and positive cultures (predominantly Pseudomonas and Klebsiella species) in 84 patients. Final diagnosis for the cause of fever in all 351 patients based on X-ray results, microbiological cultures, and clinical response to sinus drainage indicated sinusitis as the sole cause of fever in 57 (16.2%) and as contributing factor in 48 (13.8%) patients with FUO. This will underestimate the actual incidence because SXR and drainage were not performed in all patients. Conclusion Physicians treating critically ill patients should be aware of the high risk of sinusitis and take appropriate preventive measures, including the removal of nasogastric tubes in patients requiring long-term mechanical ventilation. Routine investigation of FUO should include computed tomography scan, SXR or sinus ultrasonography, and drainage should be performed if any abnormalities are found. Introduction A large proportion of patients admitted to the intensive care unit (ICU) are likely to develop fever of unknown origin (FUO) at some point of their stay there. Many of these episodes are due to well-recognised hospital-acquired infections such as ventilator-associated pneumonia (VAP) and central venous catheter infections [1,2]. Various diagnostic strategies have been developed to handle such infectious complications in the ICU, many of which have been laid down in hospital or national guidelines [3,4]. However, the potential role of sinusitis as a source of hospital-acquired infections has been much less well studied. It is well recognised that sinusitis can occur as a complication of nasotracheal intubation; however, the inci- dence of sinusitis in patients after orotracheal intubation is unclear, and the data from the literature have been conflicting [5-8]. We therefore decided to assess the role of sinusitis as CT = computed tomography; ENT = ear, nose and throat; FUO = fever of unknown origin; ICU = intensive care unit; SXR = sinus X-ray; VAP = ven- tilator-associated pneumonia. Critical Care Vol 9 No 5 van Zanten et al. R584 a hospital-acquired infection in mechanically ventilated and orotracheally intubated patients admitted to our ICU, in a pro- spective study using a rigorous protocol with predefined crite- ria for suspecting sinusitis. Our aim was not only to assess the incidence of hospital- acquired sinusitis in patients with FUO but also to provide a practical protocol for diagnostic work-up and treatment that could be quickly implemented and easily applied in everyday clinical practice. Diagnostic and therapeutic procedures were therefore chosen in part on the basis of feasibility in daily clin- ical practice in the care of critically ill patients. The three main imaging techniques available to establish a diagnosis of sinusitis are a standard sinus X-ray (SXR), ultra- sound investigation, and computed tomography (CT) of the sinuses. Of these, a CT scan of the sinus cavities is unques- tionably the most accurate and reliable procedure to establish the diagnosis of sinusitis. However, it would be highly imprac- tical and costly to perform repeated CT scans on large num- bers of ICU patients on a routine basis. In addition, transporting critically ill patients from the ICU to the radiology unit to perform a CT scan involves some risks [9-11]. A rela- tively new and promising development is the use of ultrasound as a diagnostic tool for sinusitis in the ICU setting, especially for the detection of maxillary sinusitis [12-15]; however, the reliability of this technique is strongly operator-dependent, and its sensitivity, especially in detecting frontal sinusitis, and over- all specificity are relatively low [15-20]. Varonen and associ- ates performed a meta-analysis of studies comparing SXR and ultrasound and reported that ultrasound was slightly less accurate than radiography when compared with the gold standard of sinus puncture [21]. Engels and associates [22] also concluded that, in spite of some limitations, sinus radiog- raphy rather than ultrasonography should still be viewed as the most reliable initial screening procedure for sinusitis. The most recent European Position Paper on Rhinosinusitis and Nasal Polyps recommends a combination of SXR followed by sinus puncture and aspiration as the diagnostically most accurate procedure [23]. It should be pointed out that most of these studies were not performed in mechanically ventilated ICU patients, and some studies have suggested that ultrasound has a higher sensitivity and specificity in the ICU setting. However, ultrasound has not so far been widely adopted as a first-line diagnostic tool for sinusitis, and most ICUs use plain SXRs as a first-line screen- ing tool. We therefore chose SXR as our initial screening technique. Methods Patients The study was performed in accordance with guidelines laid down by the hospital ethics committee. All mechanically venti- lated adult patients admitted to the surgical wing of our inten- sive care department during the 18-month study period who spent more than 48 hours in the ICU and who developed fever during their ICU stay were included in the study. Inclusion cri- teria were as follows: age 18 to 80 years; core temperature 38.5°C (measured in oesophagus, bladder or rectum); not admitted for infections or, if infection was the primary reason for admission, infection treated and temperature normalised for at least 72 hours before recurrence of FUO. At the time of our study, gastric tubes were inserted nasally in most patients. Sedation and analgesia were given in the context of a nurse- driven sedation protocol using the Ramsey score to guide lev- els of sedation. Exclusion criteria included severe head and facial injuries, skull fractures and immunocompromised patients. FUO was defined as follows: the cause of fever not immedi- ately clear; the patient was not admitted because of fever or sepsis, or the patient had recovered from one or more previous septic episodes or infections. This means that some patients were admitted with, for example, abdominal sepsis, and devel- oped sinusitis in the course of their admission. Such patients were eligible for inclusion in our study. Protocol According to our protocol all patients who developed fever first underwent routine analysis, which included physical examination, drawing of blood cultures and analysis for white blood cell count, and a chest X-ray. Central lines were changed if they had been in place for 1 week or more, or if there were any signs of local infection [2]. An SXR was taken if a cause of fever did not become clear from the above mentioned analysis. An SXR was also taken if a cause of fever was found on routine analysis but when fever persisted for more than 48 hours in spite of antibiotic therapy to exclude sinusitis as the primary cause of fever and/or a con- tributing factor. SXRs were taken in two directions, the straight anterior–pos- terior view (Caldwell view) and the lateral view, using portable devices in the ICU. Additional X-rays were taken if the first X- rays were difficult to interpret, in accordance with our routine for radiodiagnostic procedures [24]. Interpretations were made by the attending physician and confirmed by an inde- pendent radiologist. Three categories were used: abnormal (clouded sinuses with fluid), equivocal and normal. Patients with an abnormal SXR were treated by an ear, nose and throat (ENT) surgeon with diagnostic and therapeutic antral sinus tap and drainage [22,23]. The procedure had to be performed as soon as possible, but a maximum interval of 12 hours was allowed if there was a need for correction of coagulopathy. To prevent accidental contamination the nares were swabbed with chlorhexidine before puncture. Macro- scopic inspection of the aspirate was performed by the ENT Available online http://ccforum.com/content/9/5/R583 R585 surgeon using four categories: pus, purulent, bloody and clear. In all cases samples were taken for both aerobic and anaero- bic cultures. Cultures were performed using semi-quantitative methods (no growth, 0 colonies; +, 1 to 10 colonies; ++, 10 to 100 colonies; +++, more than 100 colonies), with ++ or +++ being regarded as positive and 0 or + as negative. Repeated drainage could be scheduled at the discretion of the attending ENT surgeon on clinical grounds. Patients with equivocal and normal results on SXR were followed up. In patients with equivocal results a repeat SXR was made 48 hours later unless the fever had resolved or another cause of fever had been found. In patients with normal SXRs no repeat was indicated except at the discretion of the attending physician. Final diagnosis for cause of fever in all 351 patients was based on blood, sputum and sinus cultures as applicable, chest X- rays and on clinical criteria (normalisation of temperature after removal of the central line, or after sinus drainage, response to antibiotic treatment, and so on). Statistical analyses were performed with Student's t-test for unpaired groups. Results are expressed as means ± SD. Sta- tistical significance was accepted at P < 0.05. Results The results are summarised in Fig. 1. During a period of 15 months a total of 351 patients met the initial inclusion criteria. In 153 patients a probable cause of fever was found on routine analysis. Therefore 198 patients met the criteria for SXR. Patient data and the results of these X-rays are shown in Table 1. On the basis of the results of the SXR, sinus drainage was first performed in those patients with evident abnormalities (n = 101). Drainage was performed in 98 of these 101 patients within 12 hours (mean 2.05 ± 5.7 hours). Twenty-four patients had been given platelets or fresh frozen plasma before the procedure. In three patients the procedure was delayed for a longer period because of the use of anticoagulants and/or Figure 1 Flowchart depicting the organisation of our study in patients with fever of unknown origin (FUO), as well as the diagnostic work-up and culture resultsFlowchart depicting the organisation of our study in patients with fever of unknown origin (FUO), as well as the diagnostic work-up and culture results. Critical Care Vol 9 No 5 van Zanten et al. R586 platelet aggregation inhibitors. In these patients the procedure was performed within 48 hours. Repeat drainage was performed in 41 patients after an aver- age period of 52 ± 38 hours. The initial (macroscopic) interpretation of the material obtained during the draining procedure by the ENT surgeon was pus in fluid from 17 of 101 patients (17%), purulent in 38 (38%), bloody in 13 (13%) and clear in 33 (33%). Culture results of initial sinus drainage are shown in Table 2. Many patients had more than one microorganism cultured from the sinus fluid. A total of 140 microorganisms were cultured from 84 of these 101 sinus drainage fluids (84%). All cultures that had been deemed as pus or purulent on macroscopic evalua- tion turned out positive for pathogenic microorganisms. How- ever, bacteria were also cultured from a substantial proportion (18 of 33 (55%)) of the fluids that had been deemed clear on microscopic inspection. The cultured pathogens are listed in Table 2. The most predominant microorganisms in the sinus fluids were Gram-negative pathogens such as Pseudomonas and Klebsiella species. Of the 28 patients with indeterminate or equivocal results on the initial SXR, a repeat SXR was performed in 25 patients. Ten (40%) now had obvious abnormalities, and drainage was performed. Of these the diagnosis of sinusitis was confirmed in 9 patients. Of the 69 patients with an initially normal SXR, sinus drainage was nevertheless performed in the subsequent 72-hour period in 12 patients on clinical grounds (n = 2), repeat SXR (n = 2) and following CT scan (n = 8). The diag- nosis was confirmed by drainage and cultures in all 12 of these patients. Thus, a total of 21 cases (22%) of microbio- logical sinusitis were subsequently found in the group of 97 patients who initially had equivocal or normal findings on SXR. On the basis of these clinical, radiological and microbiological criteria we evaluated the final diagnoses in all 351 patients with FUO initially included in our study. The results are shown in Tables 3 and 4. Discussion The results of our study demonstrate that sinusitis is a fre- quently occurring hospital-acquired infection in the ICU. Sinusitis was initially diagnosed in 84 of 351 (24%) patients with FUO, and in an additional 21 patients who had equivocal or normal findings on initial SXR, giving a total of 105 of 351 patients (30%). This underestimates the true incidence because SXRs were not taken in 153 patients who had obvi- ous other causes of fever on initial screening, some of whom might also have had sinusitis. Sinusitis was the sole cause of fever in 57 patients (16%) and one of several causes (for example sinusitis and purulent bron- chitis) in 48 patients (14%). Pathogenic microorganisms were cultured not only when material obtained by antral sinus punc- ture was classified as 'purulent' but also in more than half of the patients whose puncture material was less suspect on macroscopic examination. Previous studies on sinusitis in orotracheally intubated patients have reported a lower incidence of sinusitis than was observed in our study, ranging from 2% to 7.7% [5,7,25,26]. There are several possible reasons for this. First, the rate of antibiotic resistance in the Netherlands is low, and antibiotics are used relatively sparingly. This might have reduced the like- lihood of undetected sinusitis being concomitantly treated because patients were receiving antibiotics for other infec- tions [25,27]. Second, our patients were more severely ill than patients included in the previous studies, as demonstrated by a high average severity of disease score (Acute Physiology and Chronic Health Evaluation (APACHE)-II score of 21 ± 6.8 in our study, compared with Simplified Acute Physiology Score (SAPS) II scores of 12 ± 4.5 [5] and 11.0 ± 3.5 [7]; other studies reported no severity scores). Third, risk factors for sinusitis such as sedation and nasogastric tube feeding were present more frequently in our patients, perhaps because of the greater severity of disease. Of the positive cultures in our patients, 77% contained Gram- negative pathogens. This rate is higher than reported in previ- ous studies, in which about 50% of cultured pathogens were Table 1 Epidemiological data, results of SXR and macroscopic evaluation of sinus fluids obtained by antral puncture Parameter Value Patient data (n = 351) Sex (M:F) 193:158 Age (mean ± SD) 59 ± 21.2 APACHE II score (mean ± SD) 21 ± 6.8 ICU LOS (days) at diagnosis (mean ± SD) 5.9 ± 5.7 Results of sinus X-rays in patients with FUO (n = 198) Sinus X-ray abnormal (two directions) 101 (51%) Sinus X-ray equivocal 28 (14%) Sinus X-ray normal 69 (35%) Macroscopic evaluation of sinus fluid (n = 101) Pus 17 (17%) Purulent 38 (38%) Bloody 13 (13%) Clear 33 (33%) APACHE, Acute Physiology and Chronic Health Evaluation; FUO, fever of unknown origin; UCI LOS, length of stay in the intensive care unit. Available online http://ccforum.com/content/9/5/R583 R587 Gram-positive [8,25,28,29]. This might be explained by differ- ences in case mix, severity of illness and length of ICU stay, as well as effects of previous antibiotic treatment on the patients' microflora [25,28,29]. Our study has some limitations. The diagnosis was based on abnormal findings on SXR and positive microbiological cul- tures obtained after antral drainage. However, SXRs cannot accurately distinguish purulent sinusitis from sterile fluids, so abnormal SXRs may overestimate the incidence of sinusitis [8,25]. Moreover, even positive microbiological cultures may not prove clinically relevant sinusitis, because they may indi- cate colonisation rather than actual infection. We tried to cir- cumvent these problems by classifying only cultures with more than 10 colonies of bacteria as positive and by basing our diagnosis on a combination of radiological abnormalities, pos- itive cultures, and clinical response to therapeutic measures such as drainage and targeted antibiotic treatment. We are therefore confident that our results accurately reflect the true incidence of sinusitis. Early detection and treatment is important because delays can lead to the development of VAP, sepsis, and life-threatening complications such meningitis, mastoiditis, intra-cranial Table 2 Results of sinus fluid cultures of patients with gross abnormalities in their initial sinus X-ray Bacterium Sinus fluid Same MO cultured from tracheal aspirate Same MO cultured from blood Pseudomonas aeruginosa 32 22 8 Klebsiella oxytoca 531 Klebsiella pneumoniae 17 10 2 Enterococcus faecalis 10 2 1 Enterobacter cloacae 12 4 0 Escherichia coli 16 8 1 Staphylococcus aureus 831 Gram-positive mixture 11 - - Gram-negative mixture 14 - - Other a 15 8 1 Total 140 60 15 There were 101 patients with gross abnormalities in their initial sinus X-ray. Positive cultures were obtained in 84 patients, with 140 different types of microorganisms (MOs). Columns 3 and 4 show positive results of the same microorganisms (MOs) cultured from tracheal aspirate and blood, cultured in the period between 24 hours before and 48 hours after sinus drainage. a Other pathogens included anaerobic bacteria (such as Bacteroides sp.) and fungi (Candida sp.). Table 3 Initial diagnosis for fever of unknown origin in mechanically ventilated patients in intensive care unit Cause of fever Sole cause One of multiple causes Central line infection 43 1 Upper respiratory tract infection/pneumonia a 93 42 Sinusitis 45 39 Abdominal focus 5 3 Other b 21 Unknown 121 Total 188 86 Initial diagnosis was performed after initial screening with physical examination and chest X-ray in all 351 patients, sinus X-ray in 198 patients and sinus drainage in 98 patients; cultures were not yet available. All patients had fever and leucocytosis. a Purulent tracheobronchial aspirate with cultures positive for pathogenic microorganisms, combined with new or progressive pulmonary infiltrates on chest X-ray; b other causes of fever included meningitis, phlebitis and deep venous thrombosis. Critical Care Vol 9 No 5 van Zanten et al. R588 abscesses and venous thrombosis of the sinus cavernosus [24,30,31]. Early treatment of sinusitis may significantly reduce the risk of VAP and perhaps also ICU mortality [8,32]. The results of our microbiological cultures underline the close link between sinusitis and the development of VAP. Of 105 patients in whom positive sinus cultures were obtained, the same microorganisms were cultured from bronchotracheal aspirates in 40% of cases (n = 42). In some patients we were able to demonstrate that positive sinus cultures preceded positive cultures from the lungs, strongly suggesting that sinusitis can lead to infections of the lower airway. Others reported similar observations; for example, Holzapfel and co- workers found that the early detection and treatment of hospi- tal-acquired sinusitis could prevent the occurrence of VAP and reduce mortality in nasotracheally intubated ICU patients [33]. Bacteraemia with the same microorganism as that cultured from the sinus occurred in 12 patients; in five patients the microorganism causing bacteraemia was cultured only from the blood and the sinus, making sinusitis the most likely cause of bacteraemia. However, no definite conclusions about cause and effect can be drawn because bacteraemia can also lead to sinusitis, with bacterial colonisation of sinus fluids following bacteraemia [30,34]. Various mechanisms might explain the high incidence of sinusitis in ICU patients. The first is anatomical. The paranasal sinuses secrete mucus that flows to the natural ostia located posteriorly towards the nasopharynx; this flow can be blocked by infection, inflammation, anatomic abnormalities or the pres- ence of foreign material such as nasotracheal intubation tubes. Even tubes with smaller diameters (such as nasogastric feed- ing and suction tubes) can cause significant obstruction in the normal flow of sinus fluids, leading to an increased risk of bac- terial colonisation and development of hospital-acquired sinusitis [25,28]. The presence of nasogastric tubes has been linked to a significant increase in the risk for sinusitis in mechanically ventilated patients (odds ratio 14.1, 95% confi- dence interval 1.7 to 117) [25]. Another important risk factor is the use of sedatives (odds ratio 15.9, 95% confidence inter- val 1.9 to 133.5) [25]. Underlying mechanisms may include the suppression of normal cleansing mechanisms such as coughing, sneezing and nose-blowing, because of sedation and analgesia; in addition, immobility precludes positional changes that improve mucous drainage under normal circum- stances [24]. Remaining in a recumbent position can increase nasal congestion and obstruction of the ostia of the maxillary sinuses. This problem may be compounded by the positive inspiratory and end-expiratory pressure in ventilated patients, which also induces an increase in central venous pressure [6,35]. In addition, critically ill patients recovering from earlier episodes of sepsis may develop relative immune suppression, so-called immunoparalysis [36]. ICU patients are often unable to communicate, and complaints related to sinusitis may go unnoticed by the medical and nurs- ing staff. Patients may have a 'runny nose', or discharge of purulent material from the nasal cavity. However, this is seen in only 27% of cases [37]. Thus elevations in white blood cell count and/or FUO may be the first presenting symptoms [24]. In theory, the use of imaging modalities such as CT scans [24,38,39] and B-mode ultrasound [12-14] could improve the diagnostic yield. As discussed above, the CT scan should be regarded as the gold standard for the diagnosis of sinusitis. Unfortunately, CT scans are not easily performed in the ICU setting, meaning that the patient has to be transported to the department of radiology for this procedure. These in-hospital transports can be risky [9,10,40,41]. The potential benefits in establishing or confirming the diagnosis should therefore be weighed against the risks of transport. The development of mobile CT scans for use at bedside would significantly reduce Table 4 Final diagnosis for FUO at ICU discharge, with final results of all cultures known Cause of fever Sole cause One of multiple causes a Central line infection 44 11 Upper respiratory tract infection/pneumonia 132 58 Sinusitis 57 48 Abdominal focus 8 16 Other b 12 28 Unknown 46 Total 253 161 See also Fig. 1. a Most patients with more than one cause of fever had sinusitis and bronchitis/pneumonia; b other causes of fever included meningitis (not related to sinusitis), phlebitis and deep venous thrombosis. Available online http://ccforum.com/content/9/5/R583 R589 these problems; however, such devices are not yet available in most hospitals. Some authors have suggested that ultrasound may provide a good or even better alternative to SXR for detecting sinusitis at bedside in critically ill patients [12,13]. However, ultrasound is not yet widely used for this purpose in routine clinical prac- tice. Moreover, its diagnostic accuracy depends on the expe- rience of the operator, and the costs are higher than for SXR. In addition, the literature comparing diagnostic yields of ultra- sound and SXR provides conflicting results [12,13,21,22,42]. Our study was not designed to compare the two techniques; we based our choice mainly on the fact that ultrasound is not yet widely used to detect sinusitis in the ICU setting, and on our pre-existing clinical protocols. It seems unlikely that use of ultrasound for initial screening would have significantly affected our results; at best it could have increased our diag- nostic yield, further strengthening our observation that sinusi- tis is a frequent cause of FUO in ICU patients. In addition, about 85% of the cases of hospital-acquired sinusitis associ- ated with mechanical ventilation involve the maxillary sinuses [6]. As conventional SXRs are most reliable in detecting max- illary sinusitis (in comparison with frontal and ethmoidal sinusi- tis) we feel that SXRs remain the most practical diagnostic tool, with an acceptable sensitivity for detecting sinusitis in the ICU setting. Hospitals favouring ultrasonography as initial screening method could easily adapt our protocol, replacing SXR by ultrasound. The CT scan remains the radiological gold standard in the diagnosis of sinusitis. On the basis of the results of our study we recommend that hospital-acquired sinusitis be considered in all patients with FUO in the ICU in whom a cause of fever is not immediately apparent from initial examination and chest X-ray. SXR or ultra- sound or (if possible) a CT scan should be included in the diagnostic work-up, and sinus puncture with drainage should be performed in case of abnormal or equivocal findings. In our study all procedures were performed at thr bedside; 40% of patients with confirmed sinusitis required repeat drainage, but no patients required more than two procedures. All nasal tubes should be removed if sinusitis is suspected; antibiotics should be started empirically or based on Gram staining, and adjusted for final culture results. In most patients temperature normalises within 48 hours [37]; this was also observed in our study. Radiological signs of sinusitis clear more slowly but should disappear within ±1 week [43]. The results of our study have led to the implementation of sev- eral measures to reduce the incidence of sinusitis. First, nasogastric tubes are no longer used in intubated patients unless it is expected that the endotracheal tube can be removed within 24 hours. Gastric tubes in all other patients are now inserted through the mouth. Second, patients intubated for ≥24 hours now routinely receive topical administration of saline 0.9% and/or decongestants such as xylometazoline drops in the nasal cavities. Thirdly, the nursing staff keeps a far more rigorous watch for signs of purulent nasal discharge in all patients, and diagnostic procedures such as X-sinus are per- formed if such discharge is observed. Finally, the routine diag- nostic work-up in patients who develop fever in the ICU now includes an SXR. Drainage (both as a diagnostic and thera- peutic tool) takes place in all patients with clear or equivocal signs of sinusitis. Topical decongestants are used to reduce oedema and facilitate drainage. In patients with clear SXR in whom no other diagnosis is established, SXR is repeated after 48 hours. These measures have led to a marked reduction in the incidence of sinusitis in our ICU. Conclusion Hospital-acquired sinusitis is a frequent cause of FUO in oro- tracheally intubated and mechanically ventilated critically ill patients. ICU physicians should be aware of the numerous risk factors for sinusitis simultaneously present in ICU patients and take appropriate preventive measures. We recommend includ- ing an SXR in the routine work-up for FUO in all ICU patients; drainage should take place if SXR reveals clouding, and should also be considered if the SXR is equivocal or difficult to interpret. A normal SXR does not rule out sinusitis, and when in doubt drainage or additional diagnostic procedures such as CT scan should be performed. Competing interests The author(s) declare that they have no competing interests. Authors' contributions KHP, JMD and ARG designed and coordinated the study. AvZ, RdB, JMD and KHP were involved in the collection, sta- tistical analysis and interpretation of the data. MDN performed literature analysis and assisted in the data collection. AvZ and KHP drafted and revised the manuscript. All authors read and approved the final manuscript. Key messages • Sinusitis is a frequent cause of FUO in the ICU (in this study it was the sole cause in 16% and a contributing factor in 13% of patients with FUO). • Bacterial colonisation of the sinuses often precedes the development of bronchitis and VAP; sinusitis may be a frequent cause of hospital-acquired bronchitis and VAP. • Diagnostic work-up of FUO should include an SXR, ultrasound or CT scan; drainage should be performed if any abnormalities are found. • Physicians treating critically ill patients should be aware of the high risk of sinusitis and take appropriate preven- tive measures, including the removal of nasogastric tubes in patients requiring long-term mechanical ventilation. Critical Care Vol 9 No 5 van Zanten et al. R590 References 1. Chastre J, Fagon J: Ventilator-associated pneumonia. Am J Respir Crit Care Med 2002, 165:867-903. 2. Polderman KH, Girbes AR: Central venous catheter use. Part 2: Infectious complications. Intensive Care Med 2002, 28:18-28. 3. Anon: Hospital-acquired pneumonia in adults: diagnosis, assessment of severity, initial antimicrobial therapy, and pre- ventive strategies. A consensus statement. American Thoracic Society, November 1995. Am J Respir Crit Care Med 1996, 153:1711-1725. 4. O'Grady NP, Alexander M, Dellinger EP, Gerberding JL, Heard SO, Maki DG, Masur H, McCormick RD, Mermel LA, Pearson ML, et al.: Guidelines for the prevention of intravascular catheter- related infections. Centers for Disease Control and Prevention. MMWR Recomm Rep 2002, 51:1-29. 5. Bach A, Boehrer H, Schmidt H, Geiss HK: Nosocomial sinusitis in ventilated patients. Nasotracheal versus orotracheal intubation. Anaesthesia 1992, 47:335-339. 6. Michelson A, Schuster B, Kamp HD: Paranasal sinusitis associ- ated with nasotracheal and orotracheal long-term intubation. Arch Otolaryngol Head Neck Surg 1992, 118:937-939. 7. Salord F, Gaussorgues P, Marti-Flich J, Sirodot M, Allimant C, Lyonnet D, Robert D: Nosocomial maxillary sinusitis during mechanical ventilation: a prospective comparison of orotra- cheal versus the nasotracheal route for intubation. Intensive Care Med 1990, 16:390-393. 8. Holzapfel L, Chastang C, Demingeon G, Bohe J, Piralla B, Coupry A: A randomized study assessing the systematic search for maxillary sinusitis in nasotracheally mechanically ventilated patients. Influence of nosocomial maxillary sinusitis on the occurrence of ventilator-associated pneumonia. Am J Respir Crit Care Med 1999, 159:695-701. 9. Beckmann U, Gillies DM, Berenholtz SM, Wu AW, Pronovost P: Incidents relating to the intra-hospital transfer of critically ill patients. An analysis of the reports submitted to the Australian Incident Monitoring Study in Intensive Care. Intensive Care Med 2004, 30:1579-1585. 10. Smith I, Fleming S, Cernaianu A: Mishaps during transport from the intensive care unit. Crit Care Med 1990, 18:278-281. 11. Waydhas C: Intrahospital transport of critically ill patients. Crit Care 1999, 3:R83-R89. 12. Hilbert G, Vargas F, Valentino R, Gruson D, Chene G, Bebear C, Gbikpi-Benissan G, Cardinaud JP: Comparison of B-mode ultra- sound and computed tomography in the diagnosis of maxillary sinusitis in mechanically ventilated patients. Crit Care Med 2001, 29:1337-1342. 13. Puidupin M, Guiavarch M, Paris A, Caroff P, Boutin JP, Le Bivic T, Garcia JF: B-mode ultrasound in the diagnosis of maxillary sinusitis in intensive care unit. Intensive Care Med 1997, 23:1174-1175. 14. Lichtenstein D, Biderman P, Meziere G, Gepner A: The 'sinuso- gram', a real-time ultrasound sign of maxillary sinusitis. Inten- sive Care Med 1998, 24:1057-1061. 15. Lucchin F, Minicuci N, Ravasi MA, Cordella L, Palu M, Cetoli M, Borin P: Comparison of A-mode ultrasound and computed tomography: detection of secretion in maxillary and frontal sinuses in ventilated patients. Intensive Care Med 1996, 22:1265-1268. 16. Haapaniemi J, Laurikainen E: Ultrasound and antral lavage in the examination of maxillary sinuses. Rhinology 2001, 39:39-42. 17. Pfleiderer AG, Drake-Lee AB, Lowe D: Ultrasound of the sinuses: a worthwhile procedure? A comparison of ultrasound and radiography in predicting the findings of proof puncture on the maxillary sinuses. Clin Otolaryngol Allied Sci 1984, 9:335-339. 18. Jensen C, von Sydow C: Radiography and ultrasonography in paranasal sinusitis. Acta Radiol 1987, 28:31-34. 19. Trigaux JP, Bertrand BM, Van Beers BE: Comparison of B-mode ultrasound, radiography and sinuscopy in the diagnosis of maxillary sinusitis. Report on 177 cases. Acta Otorhinolaryngol Belg 1988, 42:670-679. 20. Shapiro GG, Furukawa CT, Pierson WE, Gilbertson E, Bierman CW: Blinded comparison of maxillary sinus radiography and ultrasound for diagnosis of sinusitis. J Allergy Clin Immunol 1986, 77:59-64. 21. Varonen H, Makela M, Savolainen S, Laara E, Hilden J: Compari- son of ultrasound, radiography, and clinical examination in the diagnosis of acute maxillary sinusitis: a systematic review. J Clin Epidemiol 2000, 53:940-948. 22. Engels EA, Terrin N, Barza M, Lau J: Meta-analysis of diagnostic tests for acute sinusitis. J Clin Epidemiol 2000, 53:852-862. 23. European Academy of Allergology and Clinical Immunology: Euro- pean position paper on rhinosinusitis and nasal polyps. Rhinol Suppl 2005, 18:1-87. 24. Seiden AM: Sinusitis in the critical care patient. New Horiz 1993, 1:261-270. 25. George DL, Falk PS, Umberto Meduri G, Leeper KV Jr, Wunderink RG, Steere EL, Nunnally FK, Beckford N, Mayhall CG: Nosoco- mial sinusitis in patients in the medical intensive care unit: a prospective epidemiological study. Clin Infect Dis 1998, 27:463-470. 26. Borman KR, Brown PM, Mezera KK, Jhaveri H: Occult fever in sur- gical intensive care unit patients is seldom caused by sinusitis. Am J Surg 1992, 164:412-415. 27. Cars O, Molstad S, Melander A: Variation in antibiotic use in the European Union. Lancet 2001, 357:1851-1853. 28. Vandenbussche T, De Moor S, Bachert C, Van Cauwenberge P: Value of antral puncture in the intensive care patient with fever of unknown origin. Laryngoscope 2000, 110:1702-1706. 29. Bert F, Lambert-Zechovsky N: Microbiology of nosocomial sinusitis in intensive care unit patients. J Infect 1995, 31:5-8. 30. Bos AP, Tibboel D, Hazebroek FW, Hoeve H, Meradji M, Molenaar JC: Sinusitis: hidden source of sepsis in postoperative pediat- ric intensive care patients. Crit Care Med 1989, 17:886-888. 31. O'Reilly MJ, Reddick EJ, Black W, Carter PL, Erhardt J, Fill W, Maughn D, Sado A, Klatt GR: Sepsis from sinusitis in nasotra- cheally intubated patients. A diagnostic dilemma. Am J Surg 1984, 147:601-604. 32. Talmor M, Li P, Barie PS: Acute paranasal sinusitis in critically ill patients: guidelines for prevention, diagnosis and treatment. Clin Infect Dis 1997, 25:1441-1446. 33. Holzapfel L, Chevret S, Madinier G, Ohen F, Demingeon G, Cou- pry A, Chaudet M: Influence of long-term oro- or nasotracheal intubation on nosocomial maxillary sinusitis and pneumonia: results of a prospective, randomized, clinical trial. Crit Care Med 1993, 21:1132-1138. 34. Aebert H, Hunefeld C, Regel G: Paranasal sinusitis and sepsis in ICU patients with nasotracheal intubation. Intensive Care Med 1988, 15:27-30. 35. Aust R, Drettner B: The patency of the maxillary ostium in rela- tion to body posture. Acta Otolaryngol 1975, 80:443-446. 36. Adrie C, Pinsky MR: The inflammatory balance in human sepsis. Intensive Care Med 2000, 26:364-375. 37. Caplan ES, Hoyt NJ: Nosocomial sinusitis. JAMA 1982, 247:639-641. 38. Osguthorpe JD, Hadley JA: Rhinosinusitis. Current concepts in evaluation and management. Med Clin North Am 1999, 83:27-41. 39. Zinreich SJ: Rhinosinusitis: radiologic diagnosis. Otolaryngol Head Neck Surg 1997, 117:S27-S34. 40. Szem JW, Hydo LJ, Fischer E, Kapur S, Klemperer J, Barie PS: High-risk intrahospital transport of critically ill patients: safety and outcome of the necessary 'road trip'. Crit Care Med 1995, 23:1660-1666. 41. Evans A, Winslow EK: Oxygen saturation and hemodynamic response in critically ill, mechanically ventilated adults during intrahospital transport. Am J Crit Care 1995, 4:106-111. 42. de Bock GH, Houwing-Duistermaat JJ, Springer MP, Kievit J, van Houwelingen JC: Sensitivity and specificity of diagnostic tests in acute maxillary sinusitis determined by maximum likelihood in the absence of an external standard. J Clin Epidemiol 1994, 47:1343-1352. 43. Deutschman CS, Wilton P, Sinow J, Dibbell D Jr, Konstantinides FN, Cerra FB: Paranasal sinusitis associated with nasotracheal intubation: a frequently unrecognized and treatable source of sepsis. Crit Care Med 1986, 14:111-114. . causes of fever on initial screening, some of whom might also have had sinusitis. Sinusitis was the sole cause of fever in 57 patients (16%) and one of several causes (for example sinusitis and. that sinusitis can occur as a complication of nasotracheal intubation; however, the inci- dence of sinusitis in patients after orotracheal intubation is unclear, and the data from the literature. drawn because bacteraemia can also lead to sinusitis, with bacterial colonisation of sinus fluids following bacteraemia [30,34]. Various mechanisms might explain the high incidence of sinusitis